Reanalysis in sentence processing: Evidence against current constraint-based and two-stage models

Two eye-tracking experiments investigated processing of VP-NP attachment am biguities. Experiment 1 tested sentences in which there was an initial bias toward VP attachment. Readers experienced more difficulty when semantic in formation disambiguated the sentences to NP attachment than when it disambi guated them to VP attach ment or when it was consistent with tither analysi s. Experiment 2 tested sentences in which there was no initial bias toward either VP or NP attachment. Readers experienced more difficulty when semant ic information disambiguated the sentences to NP attachment or VP attachmen t than when it was consistent with either analysis. We argue th;lt these re sults challenge theories that assume a competition mechanism, such as const raint-based lexicalist accounts (e.g., MacDonald, Pearlmutter, & Seidenberg , 1994; McRae, Spivey-Knowlton, & Seidenberg, 1998; Spivey-Knowlton & Sediv y, 1995) and fixed-choice two-stage models (e.g., Frazier, 1987). We interp ret the results in terms of the unrestricted race model (cf. Traxler, Picke ring, & Clifton, 1998). (C) 2001 Academic Press.